Psychiatric comorbidity is common among persons with substance abuse problems, and mental health care in conjunction with substance abuse treatment is associated with better health outcomes. It is known that racial/ethnic minorities are less likely than are whites to receive treatment for substance abuse and mental health symptoms; however, the specific reasons for these disparities are not well understood. The general aim of this study is to identify the role of community-level determinants such as provider supply, poverty, and other aspects of the social environment in explaining racial/ethnic disparities in mental health services use and addiction-related health outcomes among substance abuse treatment clients. We will analyze data from the California Treatment Outcome Project (CalTOP), a computer-based system developed to support the evaluation of publicly funded substance abuse programs in California. We will test the following hypotheses: 1. Racial and ethnic minorities are less likely than are whites to obtain mental health care and continue with outpatient treatment after discharge from substance abuse treatment. 2. Racial and ethnic disparities in mental health service use and addiction-related health outcomes are due, n large part, to differences in the availability of mental health service providers and other measures of community disorganization such as poverty and residential instability. This analysis will focus on the 1,091 clients who needed mental health services and completed the baseline, three-month, and nine-month interviews. We will use multi-level modeling to examine the role of community-level characteristics in explaining disparities in mental health service use, with data from Client and Service Information system, an administrative database managed by the California Department of Mental Health. A shortened version of the Addiction Severity Index (ASI) was administered at baseline and nine-month follow- up. We will examine racial/ethnic differences in ASI alcohol/drug and psychiatric status scores at nine-month follow-up, and use multivariate regression to examine the role of community-level characteristics in explaining disparities in addiction severity. [unreadable] [unreadable] [unreadable]